Package org.apache.hadoop.hdfs.server.blockmanagement
package org.apache.hadoop.hdfs.server.blockmanagement
-
ClassDescriptionSpace balanced block placement policy.Space balanced rack fault tolerant block placement policy.This interface is used by the block manager to expose a few characteristics of a collection of Block/BlockUnderConstruction.BlockIdManager allocates the generation stamps and the block ID.For a given block (or an erasure coding block group), BlockInfo class maintains 1) the
BlockCollectionit is part of, and 2) datanodes where the replicas of the block, or blocks belonging to the erasure coding block group, are stored.Subclass ofBlockInfo, used for a block with replication scheme.Subclass ofBlockInfo, presenting a block group in erasure coding.This class contains datanode storage information and block index in the block group.Keeps information related to the blocks stored in the Hadoop cluster.Used to inject certain faults for testing.This interface is used for choosing the desired number of targets for placing block replicas.The class is responsible for choosing the desired number of targets for placing block replicas.The class is responsible for choosing the desired number of targets for placing block replicas.The class is responsible for choosing the desired number of targets for placing block replicas on environment with node-group layer.The class is responsible for choosing the desired number of targets for placing block replicas that honors upgrade domain policy.An implementation of @see BlockPlacementStatus forAn implementation of @see BlockPlacementStatus forThis is an interface used to retrieve statistic information related to block management.A collection of block storage policies.Represents the under construction feature of a Block.Scans the namesystem, scheduling blocks to be cached as appropriate.This class manages datanode configuration using a json file.Stores information about all corrupt blocks in the File System.The corruption reason codeThis class implements the logic to track decommissioning and entering maintenance nodes, ensure all their blocks are adequately replicated before they are moved to the decommissioned or maintenance state.Checks to see if datanodes have finished DECOMMISSION_INPROGRESS or ENTERING_MAINTENANCE state.Manages decommissioning and maintenance state for DataNodes.This abstract class provides some base methods which are inherited by the DatanodeAdmin BackOff and Default Monitors, which control decommission and maintenance mode.Interface used to implement a decommission and maintenance monitor class, which is instantiated by the DatanodeAdminManager class.This class extends the DatanodeInfo class with ephemeral information (eg health, capacity, what blocks are associated with the Datanode) that is private to the Namenode, ie this class is not exposed to clients.Block and targets pairA list of CachedBlock objects on this datanode.Manage datanodes, include decommission and other activities.Datanode statisticsA Datanode has one or more storages.This interface is used for retrieving the load related statistics of the cluster.This interface abstracts how datanode configuration is managed.This class manages the include and exclude files for HDFS.The HostSet allows efficient queries on matching wildcard addresses.A immutable object that stores the number of live replicas and the number of decommissioned Replicas.This exception is thrown when the name node runs out of V1 (legacy) generation stamps.This class allows us to manage and multiplex between storages local to datanodes, and provided storage.An abstract DatanodeDescriptor to track datanodes with provided storages.Generate the next valid block group ID by incrementing the maximum block group ID allocated so far, with the first 2^10 block group IDs reserved.Generate the next valid block ID by incrementing the maximum block ID allocated so far, starting at 2^30+1.This class aggregates information fromSlowDiskReportsreceived via heartbeats.This structure is a thin wrapper over disk latencies.Disabled tracker for slow peers.This class aggregates information fromSlowPeerReportsreceived via heartbeats.Statistics per StorageType.This exception is thrown if resolving topology path for a node fails.